Plotting interaction predicting laten... PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
 Rachel posted on Thursday, March 16, 2017 - 11:51 am
Dear Dr. Muthén,

I ran a LGC model with 3 exogenous variables (2 observed variables and their product term), predicting the latent slope and latent intercept.

I have found code to plot:
- an interaction between observed variables predicting an observed variable
- an interaction between latent variables predicting a latent variable
but am unsure whether either of these would be appropriate for an interaction between observed variables predicting latent variables, particularly a latent intercept?

I am only concerned with interpreting the interaction predicting the latent intercept.

1) Where can I find the code to plot the interaction of 2 observed variables predicting a latent intercept (plotting 3 groups the Aiken & West way)?

2) What would be the code to significance test the simple slopes plotted?

3) Is there anything to be mindful of regarding interpreting an interaction predicting a latent intercept?
Or can the plot be interpreted like any other interaction between continuous variables, plotted for groups at the mean and +/- 1SD (just framed instead in terms of where subjects start/scores at Time 1, e.g., cases 1SD above the mean for X1 were associated with high Y scores at T1 regardless of level of X2, but for cases 1SD below the mean of X1, high levels of X2 were associated with high scores in Y at T1)?

Thank you!
 Bengt O. Muthen posted on Friday, March 17, 2017 - 5:57 pm
The approach for studying an interaction is the same whether the DV is latent or observed (as long as it is continuous which is the case for you).

Plotting of interactions,simple slopes, and moderation is described in our new book.
Back to top
Add Your Message Here
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Options: Enable HTML code in message
Automatically activate URLs in message